betaFeedback

WorksPhysics and Astronomy

Thermodynamic Coupling of Mass and Electromagnetic Fields: Entropic Origin of Parity Asymmetry and the Meissner Effect

Abstract

We develop a thermodynamic framework that couples mass dynamics, described by the Newton- Gibbs-van der Waals formalism, with electromagnetic fields beyond the scope of classical Maxwell theory. Classical Newtonian mechanics does not capture density evolution in the momentum balance, while the standard Maxwell equations neglect the contribution of the curl component of the electric field associated with moving charges. Building on an alternative understanding on entropy, we develop a generalized theory for electrodynamics governed by entropy-production constraints. The resulting framework yields a modified Maxwell stress tensor that incorporates the moving-charge contribution, leading to intrinsic parity asymmetry in electromagnetic forces. The theory naturally reproduces key features of superconductivity, including the Meissner effect, and reduces to the conventional Maxwell-Faraday and Maxwell-Ampere equations in an appropriate limit. This entropic formulation provides a unified thermodynamic basis for mass-field coupling and reveals new physical consequences arising from motion-induced electromagnetic effects.

Introduction

In classical physics, mass systems and electromagnetic fields are treated within distinct frameworks. The behavior of mass systems is typically described using the Newton-Gibbs-van der Waals formulation [1, 2], while the dynamics of electromagnetic fields are governed by the classical Maxwell equations [3]. Both frameworks are well established and extensively studied, and it is not the aim of this work to provide a comprehensive review of their individual developments. Instead, we focus on the coupling between the Newton-Gibbs-van der Waals framework and Maxwell’s equations. Establishing such a coupling remains an open problem, primarily due to the following challenges:

(I): Newtonian mechanics relates the force F\mathbf{F} to inertial through mam\mathbf{a} as F=ma\mathbf{F}=m\mathbf{a}, where mm is the particle mass and a=du/dt\mathbf{a}=\mathrm{d}\mathbf{u}/\mathrm{d}t depicts the acceleration which is the total time derivative of the velocity u\mathbf{u}. However, within this framework, Newton’s law does not account for the density evolution appearing in the momentum balance, d(ρu)/dt\mathrm{d}(\rho\mathbf{u})/\mathrm{d}t. In my previous work [4], this limitation was addressed by introducing a velocity entropy sus_u into the Gibbs-van der Waals formulation, yielding an entropy interpretation of acceleration

ρu⋅dudt=−Tdsudt.(1)\rho\mathbf{u}\cdot\frac{\mathrm{d}\mathbf{u}}{\mathrm{d}t}=-T\frac{\mathrm{d}s_u}{\mathrm{d}t}. \tag*{(1)}

Here, TT is the absolute temperature and ρ\rho represents the density.

(II): The Maxwell equations fail to capture the moving effect of charges. According to the Helmholtz decomposition, the electrical field E\mathbf{E} can be decomposed into a curl-free scalar-field ψ0\psi_0 (∇×∇ψ0=0\nabla\times\nabla\psi_0=0) and a divergence free vector-field ψ1\psi_1 [∇⋅(∇×ψ1)=0\nabla\cdot(\nabla\times\psi_1)=0]:

E=∇ψ0+∇×ψ1.(2)\mathbf{E}=\nabla\psi_0+\nabla\times\psi_1. \tag*{(2)}

The classic Maxwell equations only consider the electrostatic part, ∇ψ0\nabla\psi_0. Specifically, the classic Maxwell stress tensor does not consider the moving effect associated with the vector field, ψ1\psi_1.

In this work, I extend the entropy concept introduced in (I) for modifying Newtonian mechanics to the electromagnetic field in (II), and develop a generalized theory of electrodynamics. The proposed framework is built on a simple but fundamental principle of free-energy minimization or, equivalently, entropy maximization

ddt∫ΩsdΩ≥0(3)\frac{\mathrm{d}}{\mathrm{d}t}\int_{\Omega} s\mathrm{d}\Omega\ge0 \tag*{(3)}

by giving a new understanding on the entropy ss. Our theory offers several key advantages:

  • (a) It shows that the motion-induced contribution associated with ψ1\psi_1 modifies the classical Maxwell stress tensor. This generalized stress tensor leads naturally to parity asymmetry in electromagnetic forces.

  • (b) It is consistent with the Meissner effect in superconductors, predicting zero electrical resistance when the magnetic field is zero.

  • (c) In an appropriate limiting case, it reproduces the classical Maxwell equations, including the Maxwell-Faraday and Maxwell-Ampere laws [5].

  • (d) It provides a general definition for the speed of sound and light.

The remainder of the paper is organized as follows. In Section II, we introduce a general dissipation principle. Section III presents the derivation of the time derivative of the energy functional for a mass system, while Section IV addresses the time derivative of the energy functional for electromagnetic fields. In Section V, we formulate the kinetic equations based on the dissipation principle. Section VI provides a case study demonstrating the application of the present theory, and Section VII concludes the work.

Dissipation principle

Theorem. We consider a domain Ω\Omega with boundary Γ\Gamma. The domain consists of a binary mass system and electromagnetic field. We define a free energy functional as F(c):Φ↦R\mathcal{F}(c):\Phi\mapsto\mathbb{R} and a composition function c(t,x(t)):R+×R3↦Rc(t,\mathbf{x}(t)):\mathbb{R}^{+}\times\mathbb{R}^{3}\mapsto\mathbb{R}. We formulate the free energy functional as

F(c)=∫Ωf(c)dΩ=∫Ω(θ−Tsc)dΩ(4)\mathcal{F}(c)=\int_{\Omega}f(c)\mathrm{d}\Omega=\int_{\Omega}(\theta-Ts_c)\mathrm{d}\Omega \tag*{(4)}

The physical meaning of the notations is as follows. θ:R+×R3↦R\theta:\mathbb{R}^{+}\times\mathbb{R}^{3}\mapsto\mathbb{R} is the internal energy, TT is the absolute temperature, sc:R+×R3↦Rs_c:\mathbb{R}^{+}\times\mathbb{R}^{3}\mapsto\mathbb{R} is the entropy associated with the composition cc.

The total time derivative is expressed as

dc(t,x(t))dt=∂c∂t+u⋅∇c,(5)\frac{\mathrm{d}c(t,\mathbf{x}(t))}{\mathrm{d}t} = \frac{\partial c}{\partial t} + \mathbf{u}\cdot\nabla c, \tag*{(5)}

where u=∂x/∂t\mathbf{u} = \partial\mathbf{x}/\partial t.

For the functional given by Eq. (4), we state the following dissipation principle to derive the kinetic equations. For a closed system, the following free energy dissipation law

dFdt=∫ΩdfdcdcdtdΩ=∫ΩμdcdtdΩ≤0(6)\frac{\mathrm{d}\mathcal{F}}{\mathrm{d}t} = \int_{\Omega}\frac{\mathrm{d}f}{\mathrm{d}c}\frac{\mathrm{d}c}{\mathrm{d}t}\mathrm{d}\Omega= \int_{\Omega}\mu\frac{\mathrm{d}c}{\mathrm{d}t}\mathrm{d}\Omega\le0 \tag*{(6)}

implies a time evolution equation for the conserved variable cc:

dcdt=∇⋅Mc∇μ,(7)\frac{\mathrm{d}c}{\mathrm{d}t} = \nabla\cdot M_c\nabla\mu, \tag*{(7)}

and a time evolution equation for non-conserved variable c^\hat{c}:

dc^dt=−τμ,(8)\frac{\mathrm{d}\hat{c}}{\mathrm{d}t} = -\tau\mu, \tag*{(8)}

where McM_c and τ\tau are positive constants.

Proof of the dissipation principle: Substituting Eq. (7) into Eq. (6), applying integration by parts, and using the no-flux boundary condition, we obtain the free energy evolution with time as

dFdt=∫Ωdfdc(∇⋅Mc∇dfdc)dΩ(9)\frac{\mathrm{d}\mathcal{F}}{\mathrm{d}t} = \int_{\Omega}\frac{\mathrm{d}f}{\mathrm{d}c}\left(\nabla\cdot M_c\nabla\frac{\mathrm{d}f}{\mathrm{d}c}\right)\mathrm{d}\Omega \tag*{(9)}
=∫ΓdfdcMc∇dfdc⋅ndΓ−∫ΩMc(∇dfdc)2dΩ(10)= \int_{\Gamma}\frac{\mathrm{d}f}{\mathrm{d}c}M_c\nabla\frac{\mathrm{d}f}{\mathrm{d}c}\cdot\mathbf{n}\mathrm{d}\Gamma- \int_{\Omega}M_c\left(\nabla\frac{\mathrm{d}f}{\mathrm{d}c}\right)^2\mathrm{d}\Omega \tag*{(10)}
=−∫ΩMc(∇μ)2dΩ≤0.(11)= -\int_{\Omega}M_c(\nabla\mu)^2\mathrm{d}\Omega\le0. \tag*{(11)}

Substituting Eq. (8) into Eq. (6), we obtain

dFdt=−∫Ωτμ2dΩ≤0.(12)\frac{\mathrm{d}\mathcal{F}}{\mathrm{d}t} = -\int_{\Omega}\tau\mu^2\mathrm{d}\Omega\le0. \tag*{(12)}

The system energy is conserved when the free energy is compensated by a component-associated entropy term TscTs_c, namely, e=f+Tsce = f + Ts_c and

ddt∫ΩedΩ=0.(13)\frac{\mathrm{d}}{\mathrm{d}t}\int_{\Omega}e\mathrm{d}\Omega= 0. \tag*{(13)}

This equality is equivalent to

μdcdt=−Tdscdt.(14)\mu\frac{\mathrm{d}c}{\mathrm{d}t}=-T\frac{\mathrm{d}s_c}{\mathrm{d}t}. \tag*{(14)}

The left hand side is responsible for the minimization of the free energy, which is equivalent to the maximization of the entropy on the right hand side.

The above dissipation principle can be further applied for vector fields u\mathbf{u} and θ\boldsymbol{\theta}, where θ\boldsymbol{\theta} is a function of u\mathbf{u}, for example, θ=ρu\boldsymbol{\theta}=\rho\mathbf{u} denoting the momentum. For a closed system, the following energy dissipation law

dFdt=∫Ωθ⋅dudt dΩ(15)\frac{\mathrm{d}\mathcal{F}}{\mathrm{d}t}=\int_{\Omega}\boldsymbol{\theta}\cdot\frac{\mathrm{d}\mathbf{u}}{\mathrm{d}t}\,\mathrm{d}\Omega \tag*{(15)}

implies a time evolution equation for the conserved variable u\mathbf{u}

dudt=∇⋅τ1(∇θ+∇θT)+∇⋅τ2(∇⋅θ)I,(16)\frac{\mathrm{d}\mathbf{u}}{\mathrm{d}t}=\nabla\cdot\tau_1(\nabla\boldsymbol{\theta}+\nabla\boldsymbol{\theta}^{\mathrm{T}})+\nabla\cdot\tau_2(\nabla\cdot\boldsymbol{\theta})\mathbf{I}, \tag*{(16)}

and a time evolution equation for non-conserved variable u^\hat{\mathbf{u}}

du^dt=−τ0θ,(17)\frac{\mathrm{d}\hat{\mathbf{u}}}{\mathrm{d}t}=-\tau_0\boldsymbol{\theta}, \tag*{(17)}

where τ0\tau_0, τ1\tau_1, and τ2\tau_2 are positive materials dependent mobilities.

Proof of the energy dissipation for vectors: Substituting Eq. (16) into Eq. (15), applying integration by parts, and using the no-flux boundary condition, we obtain the energy dissipation as

dFdt=−∫Ωτ1∇θ:∇θ+τ2(∇⋅θ)2 dΩ≤0.(18)\frac{\mathrm{d}\mathcal{F}}{\mathrm{d}t}=-\int_{\Omega}\tau_1\nabla\boldsymbol{\theta}:\nabla\boldsymbol{\theta}+\tau_2(\nabla\cdot\boldsymbol{\theta})^2\,\mathrm{d}\Omega\le0. \tag*{(18)}

Substituting Eq. (17) into Eq. (15), we obtain

dFdt=−∫Ωτ0θ2 dΩ≤0.(19)\frac{\mathrm{d}\mathcal{F}}{\mathrm{d}t}=-\int_{\Omega}\tau_0\boldsymbol{\theta}^2\,\mathrm{d}\Omega\le0. \tag*{(19)}

The same understanding for the minimization of the free energy and the maximization of the entropy can be obtained when compensating the free energy by an associated entropy, TsuTs_{\mathbf{u}}.

θ⋅dudt=−Tdsudt.(20)\boldsymbol{\theta}\cdot\frac{\mathrm{d}\mathbf{u}}{\mathrm{d}t}=-T\frac{\mathrm{d}s_{\mathbf{u}}}{\mathrm{d}t}. \tag*{(20)}

Time derivative of the mass energy functional

We formulate the free energy functional of the mass system as

E=Ef+Ep+Ek.(21)E = E_f + E_p + E_k. \tag*{(21)}

The individual three contributions are listed as follows:

Chemical free energy:Ef=∫Ωf(c,∇c) dΩ,(22)\text{Chemical free energy:}\quad E_f = \int_{\Omega} f(c,\nabla c)\,\mathrm{d}\Omega, \tag*{(22)}
Pressure energy:Ep=∫Ωp1 dΩ,(23)\text{Pressure energy:}\quad E_p = \int_{\Omega} p_1\,\mathrm{d}\Omega, \tag*{(23)}
Kinetic energy:Ek=∫Ωρu⋅u dΩ.(24)\text{Kinetic energy:}\quad E_k = \int_{\Omega} \rho\mathbf{u}\cdot\mathbf{u}\,\mathrm{d}\Omega. \tag*{(24)}

Based on the van der Waals-Gibbs-Cahn theory [2, 6, 7, 8], we write the chemical free energy functional as

Ef=∫Ωf(c,∇c) dΩ=∫Ω[f0(c)+12κ(∇c)2]dΩ,(25)E_f = \int_{\Omega} f(c,\nabla c)\,\mathrm{d}\Omega= \int_{\Omega} \left[f_0(c) + \frac{1}{2}\kappa(\nabla c)^2\right]\mathrm{d}\Omega, \tag*{(25)}

where f0f_0 is the bulk free energy and κ\kappa represents the gradient energy coefficient.

By decomposing the entropy ss into thermal entropy sTs_T, mixing entropy scs_c, and velocity entropy sus_u, we express the free energy density as

f=θ0+μc−Tsϕ−Tsu−TsT.(26)f = \theta_0 + \mu c - T s_\phi- T s_u - T s_T. \tag*{(26)}

The first part is the internal energy solely determined by the temperature. The second part is the chemical contribution. The last three terms depict the entropy.

Next, we calculate the time derivative of the system energy functional to derive the kinetic equations. The time derivative of the chemical free energy functional reads

dEfdt=∫Ω∂f∂cdcdt+∂f∂∇c⋅d∇cdt dΩ.(27)\frac{\mathrm{d}E_f}{\mathrm{d}t} = \int_{\Omega} \frac{\partial f}{\partial c}\frac{\mathrm{d}c}{\mathrm{d}t} + \frac{\partial f}{\partial\nabla c}\cdot\frac{\mathrm{d}\nabla c}{\mathrm{d}t}\,\mathrm{d}\Omega. \tag*{(27)}

The total time derivative of ∇c\nabla c reads

d∇cdt=∂∇c∂t+u⋅∇∇c=∇∂c∂t+u⋅∇∇c,(28)\frac{\mathrm{d}\nabla c}{\mathrm{d}t} = \frac{\partial\nabla c}{\partial t} + \mathbf{u}\cdot\nabla\nabla c = \nabla\frac{\partial c}{\partial t} + \mathbf{u}\cdot\nabla\nabla c, \tag*{(28)}

where ∇∇c\nabla\nabla c is the covariant derivative, similar to the Hessian matrix. Thus, the second term in Equation (27) is rewritten as

∫Ω∂f∂∇c⋅d∇cdt dΩ=∫Ω∂f∂∇c⋅(∇∂c∂t)+∂f∂∇c⋅(u⋅∇∇c) dΩ.(29)\int_{\Omega}\frac{\partial f}{\partial\nabla c}\cdot\frac{\mathrm{d}\nabla c}{\mathrm{d}t}\,\mathrm{d}\Omega = \int_{\Omega}\frac{\partial f}{\partial\nabla c}\cdot\left(\nabla\frac{\partial c}{\partial t}\right) + \frac{\partial f}{\partial\nabla c}\cdot\left(\mathbf{u}\cdot\nabla\nabla c\right)\,\mathrm{d}\Omega. \tag*{(29)}

By using integration by parts and using no-flux boundary condition, the first term in Equation (29) is rewritten as

∫Ω∂f∂∇c⋅∇∂c∂t dΩ=∫Ω−(∇⋅∂f∂∇c)(dcdt−u⋅∇c) dΩ.(30)\int_{\Omega} \frac{\partial f}{\partial\nabla c} \cdot\nabla\frac{\partial c}{\partial t} \, \mathrm{d}\Omega= \int_{\Omega} -\left(\nabla\cdot\frac{\partial f}{\partial\nabla c}\right)\left(\frac{\mathrm{d}c}{\mathrm{d}t} - \mathbf{u} \cdot\nabla c\right) \, \mathrm{d}\Omega. \tag*{(30)}

The second term in Equation (29) is written as

∫Ω∂f∂∇c⋅(u⋅∇∇c) dΩ=∫Ω[(∇⋅∂f∂∇c)⋅∇]∇c⋅u dΩ.(31)\int_{\Omega} \frac{\partial f}{\partial\nabla c} \cdot\left(\mathbf{u} \cdot\nabla\nabla c\right) \, \mathrm{d}\Omega= \int_{\Omega} \left[\left(\nabla\cdot\frac{\partial f}{\partial\nabla c}\right) \cdot\nabla\right]\nabla c \cdot\mathbf{u} \, \mathrm{d}\Omega. \tag*{(31)}

Summarizing the above derivations and using the following vector calculus:

∇⋅(a⊗b)=(∇⋅a)b+(a⋅∇)b,\nabla\cdot(\mathbf{a} \otimes\mathbf{b}) = (\nabla\cdot\mathbf{a})\mathbf{b} + (\mathbf{a} \cdot\nabla)\mathbf{b},

with a=∂f∂∇c=κ∇c\mathbf{a} = \frac{\partial f}{\partial\nabla c} = \kappa\nabla c and b=∇c\mathbf{b} = \nabla c, we obtain the final expression for the time derivative of the chemical free energy functional EfE_f as

dEfdt=∫Ω[(∂f∂c−∇⋅∂f∂∇c)dcdt+∇⋅(∂f∂∇c⊗∇c)⋅u] dΩ(32)\frac{\mathrm{d}E_f}{\mathrm{d}t} = \int_{\Omega} \left[\left(\frac{\partial f}{\partial c} - \nabla\cdot\frac{\partial f}{\partial\nabla c}\right)\frac{\mathrm{d}c}{\mathrm{d}t} + \nabla\cdot\left(\frac{\partial f}{\partial\nabla c} \otimes\nabla c\right) \cdot\mathbf{u}\right] \, \mathrm{d}\Omega \tag*{(32)}
=∫Ω{μdcdt+[∇⋅(κ∇c⊗∇c)]⋅u} dΩ,(33)= \int_{\Omega} \left\{\mu\frac{\mathrm{d}c}{\mathrm{d}t} + [\nabla\cdot(\kappa\nabla c \otimes\nabla c)] \cdot\mathbf{u}\right\} \, \mathrm{d}\Omega, \tag*{(33)}

where, according to Eq. (32), the chemical potential μ\mu is defined as

μ=∂f∂c−∇⋅∂f∂∇c.(34)\mu= \frac{\partial f}{\partial c} - \nabla\cdot\frac{\partial f}{\partial\nabla c}. \tag*{(34)}

The time derivative of the mass pressure energy reads

dEpdt=∫Ωdp1dr⋅drdt dΩ=∫Ω∇p1⋅u dΩ.(35)\frac{\mathrm{d}E_p}{\mathrm{d}t} = \int_{\Omega} \frac{\mathrm{d}p_1}{\mathrm{d}\mathbf{r}} \cdot\frac{\mathrm{d}\mathbf{r}}{\mathrm{d}t} \, \mathrm{d}\Omega= \int_{\Omega} \nabla p_1 \cdot\mathbf{u} \, \mathrm{d}\Omega. \tag*{(35)}

The time derivative of the mass kinetic energy reads

dEkdt=∫Ω[d(ρu)dt⋅u+(ρu)⋅dudt] dΩ.(36)\frac{\mathrm{d}E_k}{\mathrm{d}t} = \int_{\Omega} \left[\frac{\mathrm{d}(\rho\mathbf{u})}{\mathrm{d}t} \cdot\mathbf{u} + (\rho\mathbf{u}) \cdot\frac{\mathrm{d}\mathbf{u}}{\mathrm{d}t}\right] \, \mathrm{d}\Omega. \tag*{(36)}

Time derivative of the energy functional of electromagnetic fields

The electromagnetic potential energy is integrated over the electromagnetic free energy density fbf_b as

Eb=∫Ωfb dΩ=∫Ω12(D⋅E+B⋅H) dΩ,(37)E_b = \int_{\Omega} f_b \, \mathrm{d}\Omega= \int_{\Omega} \frac{1}{2}(\mathbf{D} \cdot\mathbf{E} + \mathbf{B} \cdot\mathbf{H}) \, \mathrm{d}\Omega, \tag*{(37)}

where E\mathbf{E} and H\mathbf{H} are the displacements of the electrical and magnetic fields, respectively; D\mathbf{D} and B\mathbf{B} are the associated forces following the constitutive relations:

D=εE, B=χH.(38)\mathbf{D}=\varepsilon\mathbf{E},\ \mathbf{B}=\chi\mathbf{H}. \tag*{(38)}

where ε\varepsilon is electrical permittivity and χ\chi denotes magnetic permeability.

An important modification here is the general formulation for the displacements of the electrical field E\mathbf{E} and the magnetic field H\mathbf{H}:

E=∇ψ0+∇×ψ1; H=∇φ0+∇×φ1.(39)\mathbf{E}=\nabla\psi_0+\nabla\times\boldsymbol{\psi}_1;\ \mathbf{H}=\nabla\varphi_0+\nabla\times\boldsymbol{\varphi}_1. \tag*{(39)}

Here, ψ0\psi_0 and φ0\varphi_0 refer to the static potentials of electrical and magnetic fields. Noteworthily, we introduce two new vectors, ψ1\boldsymbol{\psi}_1 and φ1\boldsymbol{\varphi}_1 to depict the moving effect of the electromagnetic fields that have not been considered in the classic Maxwell equations. We will see these new vector terms lead to broken symmetry of parity. When E\mathbf{E} and H\mathbf{H} are continuous functions in the space, the validity of Equation (39) is ensured by the Helmholtz decomposition theory [9].

The electromagnetic free energy Eb\mathcal{E}_b plus the kinetic energy Ek\mathcal{E}_k and pressure energy Ep\mathcal{E}_p is expressed as

E=Eb+Ep+Ek,(40)\mathcal{E}=\mathcal{E}_b+\mathcal{E}_p+\mathcal{E}_k, \tag*{(40)}

with the following expressions for each part:

Eb=Ee1+Ee2, Ee1=∫Ω12(D⋅E)dΩ, Ee2=∫Ω12(B⋅H)dΩ,(41)\mathcal{E}_b=\mathcal{E}_{e1}+\mathcal{E}_{e2},\ \mathcal{E}_{e1}=\int_{\Omega}\frac{1}{2}(\mathbf{D}\cdot\mathbf{E})\mathrm{d}\Omega,\ \mathcal{E}_{e2}=\int_{\Omega}\frac{1}{2}(\mathbf{B}\cdot\mathbf{H})\mathrm{d}\Omega, \tag*{(41)}
Ek=∫Ω(B×D)⋅vdΩ, Ep=∫Ωp2dΩ.(42)\mathcal{E}_k=\int_{\Omega}(\mathbf{B}\times\mathbf{D})\cdot\mathbf{v}\mathrm{d}\Omega,\ \mathcal{E}_p=\int_{\Omega}p_2\mathrm{d}\Omega. \tag*{(42)}

The formulation of the kinetic energy is motivated by the Poynting vector. Here, we distinguish the movement velocity of the mass, u\mathbf{u}, from the propagation speed of the electromagnetic field, v\mathbf{v}, unlike assuming u=v\mathbf{u}=\mathbf{v} in Ref. [10].

Next, we derive the total time derivative of the system energy as

dEe1dt=∫Ω12ε′E2dcdt+D⋅d∇ψ0dt+D⋅d∇×ψ1dt.(43)\frac{\mathrm{d}\mathcal{E}_{e1}}{\mathrm{d}t}=\int_{\Omega}\frac{1}{2}\varepsilon'\mathbf{E}^{2}\frac{\mathrm{d}c}{\mathrm{d}t}+\mathbf{D}\cdot\frac{\mathrm{d}\nabla\psi_0}{\mathrm{d}t}+\mathbf{D}\cdot\frac{\mathrm{d}\nabla\times\boldsymbol{\psi}_1}{\mathrm{d}t}. \tag*{(43)}

The first term in Equation (43) can be adsorbed into the mass effect term [Eq. (34)], leading to a modified chemical potential plus 12ε′E2\frac{1}{2}\varepsilon'\mathbf{E}^{2}.

Applying the divergence theory to the second term in Equation (43) and no-flux boundary condition leads to

∫ΩD⋅d∇ψ0dt=∫Ω{−(∇⋅D)dψ0dt+[∇⋅(D⊗∇ψ0)]⋅v}dΩ.(44)\int_{\Omega} \mathbf{D} \cdot\frac{\mathrm{d}\nabla\psi_0}{\mathrm{d}t} = \int_{\Omega} \left\{-(\nabla\cdot\mathbf{D})\frac{\mathrm{d}\psi_0}{\mathrm{d}t} + [\nabla\cdot(\mathbf{D} \otimes\nabla\psi_0)] \cdot\mathbf{v}\right\}\mathrm{d}\Omega. \tag*{(44)}

Applying the divergence theory to the third term in Equation (43) and no-flux boundary condition results in

∫ΩD⋅d∇×ψ1dt=∫Ω{(∇×D)⋅∂ψ1∂t+∇⋅[D⊗(∇×ψ1)]⋅v−(∇⋅D)(∇×ψ1)⋅v}dΩ.(45)\int_{\Omega} \mathbf{D} \cdot\frac{\mathrm{d}\nabla\times\psi_1}{\mathrm{d}t} = \int_{\Omega} \left\{(\nabla\times\mathbf{D}) \cdot\frac{\partial\psi_1}{\partial t} + \nabla\cdot[\mathbf{D} \otimes(\nabla\times\psi_1)] \cdot\mathbf{v} - (\nabla\cdot\mathbf{D})(\nabla\times\psi_1) \cdot\mathbf{v}\right\}\mathrm{d}\Omega. \tag*{(45)}

Therefore, we obtain the final expression for the time derivative of the electrical free energy as

dEe1dt=∫Ω{−(∇⋅D)dψ0dt+(∇×D)⋅∂ψ1∂t+[∇⋅(D⊗E)]⋅v−(∇⋅D)(∇×ψ1)⋅v}dΩ.(46)\frac{\mathrm{d}\mathcal{E}_{e_1}}{\mathrm{d}t} = \int_{\Omega} \left\{-(\nabla\cdot\mathbf{D})\frac{\mathrm{d}\psi_0}{\mathrm{d}t} + (\nabla\times\mathbf{D}) \cdot\frac{\partial\psi_1}{\partial t} + [\nabla\cdot(\mathbf{D} \otimes\mathbf{E})] \cdot\mathbf{v} - (\nabla\cdot\mathbf{D})(\nabla\times\psi_1) \cdot\mathbf{v}\right\}\mathrm{d}\Omega. \tag*{(46)}

Applying the same derivation to the magnetic part results in

dEe2dt=∫Ω12χ′H2dcdt+B⋅d∇φ0dt+B⋅d∇×φ1dt.(47)\frac{\mathrm{d}\mathcal{E}_{e_2}}{\mathrm{d}t} = \int_{\Omega} \frac{1}{2}\chi'\mathbf{H}^{2}\frac{\mathrm{d}c}{\mathrm{d}t} + \mathbf{B} \cdot\frac{\mathrm{d}\nabla\varphi_0}{\mathrm{d}t} + \mathbf{B} \cdot\frac{\mathrm{d}\nabla\times\varphi_1}{\mathrm{d}t}. \tag*{(47)}

which gives rise to a formulation quite similar to Eq. (46) as

dEe2dt=∫Ω{−(∇⋅B)dφ0dt+(∇×B)⋅∂φ1∂t+[∇⋅(B⊗H)]⋅v−(∇⋅B)(∇×φ1)⋅v}dΩ.(48)\frac{\mathrm{d}\mathcal{E}_{e_2}}{\mathrm{d}t} = \int_{\Omega} \left\{-(\nabla\cdot\mathbf{B})\frac{\mathrm{d}\varphi_0}{\mathrm{d}t} + (\nabla\times\mathbf{B}) \cdot\frac{\partial\varphi_1}{\partial t} + [\nabla\cdot(\mathbf{B} \otimes\mathbf{H})] \cdot\mathbf{v} - (\nabla\cdot\mathbf{B})(\nabla\times\varphi_1) \cdot\mathbf{v}\right\}\mathrm{d}\Omega. \tag*{(48)}

As ∇⋅B=0\nabla\cdot\mathbf{B} = 0 (see Section VA), the moving effect of the magnetic field ∇×φ1\nabla\times\varphi_1 vanishes and the last term in Eq. (48) is zero. However, as ∇⋅D≠0\nabla\cdot\mathbf{D} \ne0 in media with charges, the last term in Eq. (46) does not vanish; this term depicts the movement effect of the electrical field, which is the key reason for parity asymmetry (see Section VI E).

The time derivative of the electromagnetic pressure energy reads

dEpdt=∫Ωdp2dr⋅drdtdΩ=∫Ω∇p2⋅vdΩ.(49)\frac{\mathrm{d}\mathcal{E}_p}{\mathrm{d}t} = \int_{\Omega} \frac{\mathrm{d}p_2}{\mathrm{d}r} \cdot\frac{\mathrm{d}r}{\mathrm{d}t}\mathrm{d}\Omega= \int_{\Omega} \nabla p_2 \cdot\mathbf{v}\mathrm{d}\Omega. \tag*{(49)}

The time derivative of the kinetic energy of the electric-magnetic field reads

dEkdt=∫Ω[d(B×D)dt⋅v+(B×D)⋅dvdt]dΩ.(50)\frac{\mathrm{d}\mathcal{E}_k}{\mathrm{d}t} = \int_{\Omega} \left[\frac{\mathrm{d}(\mathbf{B} \times\mathbf{D})}{\mathrm{d}t} \cdot\mathbf{v} + (\mathbf{B} \times\mathbf{D}) \cdot\frac{\mathrm{d}\mathbf{v}}{\mathrm{d}t}\right]\mathrm{d}\Omega. \tag*{(50)}

The kinetic equations

For a coupled system of mass and electromagnetic field, the total free energy is expressed as

E=Ef+Ep+Ek+Eb+Ep+Ek.(51)\mathcal{E} = \mathcal{E}_f + \mathcal{E}_p + \mathcal{E}_k + \mathcal{E}_b + \mathcal{E}_p + \mathcal{E}_k. \tag*{(51)}

For a closed system, the total free energy dissipates with time

dEdt≤0.(52)\frac{\mathrm{d}\mathcal{E}}{\mathrm{d}t} \le0. \tag*{(52)}

The total energy is conserved when an entropy contribution is included.

E=E+T∫Ω(sT+sc+su+sψ0+sψ1+sφ0+sφ1+sv) dΩ,(53)E = \mathcal{E} + T \int_{\Omega} (s_T + s_c + s_u + s_{\psi_0} + s_{\psi_1} + s_{\varphi_0} + s_{\varphi_1} + s_v)\,\mathrm{d}\Omega, \tag*{(53)}

which leads to different types of kinetic equations

μdcdt=−Tdscdt,(54)\mu\frac{\mathrm{d}c}{\mathrm{d}t} = -T\frac{\mathrm{d}s_c}{\mathrm{d}t}, \tag*{(54)}
ρu⋅dudt=−Tdsudt,(55)\rho\mathbf{u}\cdot\frac{\mathrm{d}\mathbf{u}}{\mathrm{d}t} = -T\frac{\mathrm{d}s_u}{\mathrm{d}t}, \tag*{(55)}
(B×D)⋅dvdt=−Tdsvdt,(56)(\mathbf{B}\times\mathbf{D})\cdot\frac{\mathrm{d}\mathbf{v}}{\mathrm{d}t} = -T\frac{\mathrm{d}s_v}{\mathrm{d}t}, \tag*{(56)}
(∇⋅D)dψ0dt=−Tdsψ0dt,(57)(\nabla\cdot\mathbf{D})\frac{\mathrm{d}\psi_0}{\mathrm{d}t} = -T\frac{\mathrm{d}s_{\psi_0}}{\mathrm{d}t}, \tag*{(57)}
(∇⋅B)dφ0dt=−Tdsφ0dt,(58)(\nabla\cdot\mathbf{B})\frac{\mathrm{d}\varphi_0}{\mathrm{d}t} = -T\frac{\mathrm{d}s_{\varphi_0}}{\mathrm{d}t}, \tag*{(58)}
(∇×D)⋅∂ψ1∂t=−T∂sψ1∂t,(59)(\nabla\times\mathbf{D})\cdot\frac{\partial\psi_1}{\partial t} = -T\frac{\partial s_{\psi_1}}{\partial t}, \tag*{(59)}
(∇×B)⋅∂φ1∂t=−T∂sφ1∂t.(60)(\nabla\times\mathbf{B})\cdot\frac{\partial\varphi_1}{\partial t} = -T\frac{\partial s_{\varphi_1}}{\partial t}. \tag*{(60)}

Evolution equations

According to the dissipation principle, we obtain the mass diffusion equations based on Eq. (54) as

dcdt=∇⋅Mc∇μ.(61)\frac{\mathrm{d}c}{\mathrm{d}t} = \nabla\cdot M_c\nabla\mu. \tag*{(61)}

According to Eq. (55)-Eq. (56), the kinetic equations for the velocities u\mathbf{u} and v\mathbf{v} read

dudt=−τu(ρu),(62)\frac{\mathrm{d}\mathbf{u}}{\mathrm{d}t} = -\tau_u(\rho\mathbf{u}), \tag*{(62)}
dvdt=−τv(B×D).(63)\frac{\mathrm{d}\mathbf{v}}{\mathrm{d}t} = -\tau_v(\mathbf{B}\times\mathbf{D}). \tag*{(63)}

or in the conserved way

dudt=∇⋅τu∇(ρu)(64)\frac{\mathrm{d}\mathbf{u}}{\mathrm{d}t} = \nabla\cdot\tau_u \nabla(\rho\mathbf{u}) \tag*{(64)}
dvdt=∇⋅τv∇(B×D)(65)\frac{\mathrm{d}\mathbf{v}}{\mathrm{d}t} = \nabla\cdot\tau_v \nabla(\mathbf{B} \times\mathbf{D}) \tag*{(65)}

Eq. (57)-Eq. (60) indicate the following electromagnetic equilibrium

∇⋅B=0(66)\nabla\cdot\mathbf{B} = 0 \tag*{(66)}
∇⋅D=−ρe(67)\nabla\cdot\mathbf{D} = -\rho_e \tag*{(67)}
∇×B=0(68)\nabla\times\mathbf{B} = 0 \tag*{(68)}
∇×D=0(69)\nabla\times\mathbf{D} = 0 \tag*{(69)}

Here, we have assumed that the time scale for the entropy relaxation of sψ0s_{\psi0}, sφ0s_{\varphi0}, sψ1s_{\psi1}, and sφ1s_{\varphi1} is much smaller than that of other entropies. The parameter ρe\rho_e depicts the charge density.

Momentum conservation of mass and electromagnetic field

Collecting all the u\mathbf{u} terms, we obtain the momentum conservation for mass as

d(ρu)dt=−∇p1−∇⋅(κ∇c⊗∇c)(70)\frac{\mathrm{d}(\rho\mathbf{u})}{\mathrm{d}t} = -\nabla p_1 - \nabla\cdot\left(\kappa\nabla c \otimes\nabla c\right) \tag*{(70)}

where the pressure p1p_1 is expressed by integrating the Gibbs-Duhem relation as [11]

−p1=f−μc(71)-p_1 = f - \mu c \tag*{(71)}

Collecting all the v\mathbf{v} terms, we obtain the momentum conservation for electromagnetic field as

d(B×D)dt=−∇⋅Γ+ρe(∇×ψ1)(72)\frac{\mathrm{d}(\mathbf{B} \times\mathbf{D})}{\mathrm{d}t} = -\nabla\cdot\mathbf{\Gamma} + \rho_e(\nabla\times\boldsymbol{\psi}_1) \tag*{(72)}

where the Maxwell stress tensor reads

Γ=−(12D⋅E+12B⋅H)I+D⊗E+B⊗H(73)\mathbf{\Gamma} = -\left(\frac{1}{2}\mathbf{D} \cdot\mathbf{E} + \frac{1}{2}\mathbf{B} \cdot\mathbf{H}\right)\mathbf{I} + \mathbf{D} \otimes\mathbf{E} + \mathbf{B} \otimes\mathbf{H} \tag*{(73)}

Case study, validation, and application

Replication of the classic Maxwell equations

We decompose the Maxwell stress tensor Γ\mathbf{\Gamma} into two parts:

Γ=T1+T2,(74)\mathbf{\Gamma} = \mathcal{T}_1 + \mathcal{T}_2, \tag*{(74)}

with the electric part T1\mathcal{T}_1 and magnetic part T2\mathcal{T}_2 reading

T1=−12D⋅E+D⊗E,(75)\mathcal{T}_1 = -\frac{1}{2}\mathbf{D}\cdot\mathbf{E}+\mathbf{D}\otimes\mathbf{E}, \tag*{(75)}
T2=−12B⋅H+B⊗H.(76)\mathcal{T}_2 = -\frac{1}{2}\mathbf{B}\cdot\mathbf{H}+\mathbf{B}\otimes\mathbf{H}. \tag*{(76)}

By using the vector calculus ∇⋅(a⊗b)\nabla\cdot(\mathbf{a}\otimes\mathbf{b}), we calculate the divergence of the Maxwell-stress tensor T1\mathcal{T}_1 as

∇⋅T1=−12E2∇ε+(∇⋅D)E−D×(∇×E).(77)\nabla\cdot\mathcal{T}_1=-\frac{1}{2}E^2\nabla\varepsilon+(\nabla\cdot\mathbf{D})\mathbf{E}-\mathbf{D}\times(\nabla\times\mathbf{E}). \tag*{(77)}

The divergence of the Maxwell stress tensor T2\mathcal{T}_2 reads

∇⋅T2=−12(∇χ)H2+(∇⋅B)H−B×(∇×H)+J×B+B×J,(78)\nabla\cdot\mathcal{T}_2=-\frac{1}{2}(\nabla\chi)H^2+(\nabla\cdot\mathbf{B})\mathbf{H}-\mathbf{B}\times(\nabla\times\mathbf{H})+\mathbf{J}\times\mathbf{B}+\mathbf{B}\times\mathbf{J}, \tag*{(78)}

where we have added a zero term, J×B+B×J=0\mathbf{J}\times\mathbf{B}+\mathbf{B}\times\mathbf{J}=0; J\mathbf{J} is the electrical current.

The total time derivative of the electromagnetic momentum reads

d(B×D)dt=B×dDdt−D×dBdt.(79)\frac{\mathrm{d}(\mathbf{B}\times\mathbf{D})}{\mathrm{d}t}=\mathbf{B}\times\frac{\mathrm{d}\mathbf{D}}{\mathrm{d}t}-\mathbf{D}\times\frac{\mathrm{d}\mathbf{B}}{\mathrm{d}t}. \tag*{(79)}

Collecting all the B×\mathbf{B}\times terms in Equation (79) and Equation (78), we obtain the Maxwell-Ampere equation as

dDdt=∇×H−J.(80)\frac{\mathrm{d}\mathbf{D}}{\mathrm{d}t}=\nabla\times\mathbf{H}-\mathbf{J}. \tag*{(80)}

Collecting all the D×\mathbf{D}\times terms in Equation (79) and Equation (77), we obtain the Maxwell-Faraday equation as

dBdt=−∇×E.(81)\frac{\mathrm{d}\mathbf{B}}{\mathrm{d}t}=-\nabla\times\mathbf{E}. \tag*{(81)}

The rest terms in Equation (78) and Equation (77) are

Π=[∇⋅(εE)]E+J×B−12E2∇ε−12H2∇χ=ρeE+J×B−12E2∇ε−12H2∇χ=0.(82)\begin{aligned} \boldsymbol{\Pi} &= [\nabla\cdot(\varepsilon\mathbf{E})]\mathbf{E}+\mathbf{J}\times\mathbf{B}-\frac{1}{2}E^2\nabla\varepsilon-\frac{1}{2}H^2\nabla\chi\\ &=\rho_e\mathbf{E}+\mathbf{J}\times\mathbf{B}-\frac{1}{2}E^2\nabla\varepsilon-\frac{1}{2}H^2\nabla\chi=0. \tag*{(82)} \end{aligned}

This vector term Π\boldsymbol{\Pi} must be a zero vector to ensure the momentum conservation of electromagnetic field.

Gauss's law

Performing a divergence operator to both sides of Eq. (80), we obtain the following steady state equation

v⋅∇(∇⋅D)+u⋅∇ρe=0,(83)\mathbf{v} \cdot\nabla(\nabla\cdot\mathbf{D}) + \mathbf{u} \cdot\nabla\rho_e = 0, \tag*{(83)}

where we have used the equality ∇⋅u=0\nabla\cdot\mathbf{u} = 0. When v=u\mathbf{v} = \mathbf{u}, we obtain the Gauss’s law, ∇⋅D=−ρe\nabla\cdot\mathbf{D} = -\rho_e for the electrical field, which is consistent with Eq. (67). Similarly, performing a divergence operator to both sides of Eq. (81), we derive the Gauss’s law, ∇⋅B=0\nabla\cdot\mathbf{B} = 0 for the magnetic field.

Meissner effect

Recalling the above derivation and neglecting the term ρe(∇×ψ1)\rho_e(\nabla\times\boldsymbol{\psi}_1) in Eq. (72), we see that the electromagnetic momentum evolution d(B×D)/dt=−∇⋅Γ\mathrm{d}(\mathbf{B} \times\mathbf{D})/\mathrm{d}t = -\nabla\cdot\boldsymbol{\Gamma} leads to two subsequent effects:

  • (I) Maxwell-Faraday and Maxwell-Ampere equations, which are responsible for the exchange of B\mathbf{B} and E\mathbf{E};

  • (II) A force term Π\boldsymbol{\Pi} given by Eq. (82).

To ensure the momentum conservation of electromagnetic fields and to obtain the Maxwell-Faraday and Maxwell-Ampere equations, the following equality

ρeE+J×B−12E2∇ε−12H2∇χ=0(84)\rho_e \mathbf{E} + \mathbf{J} \times\mathbf{B} - \frac{1}{2}\mathbf{E}^{2}\nabla\varepsilon- \frac{1}{2}\mathbf{H}^{2}\nabla\chi= 0 \tag*{(84)}

must hold everywhere in the domain. In vacuum, there is no charge and no electric current, i.e., ρe=0\rho_e = 0 and J=0\mathbf{J} = 0. Therefore, Eq. (84) is automatically fulfilled in vacuum. In media at the superconducting state, there is almost no dissipation of the velocities u\mathbf{u} and v\mathbf{v}. The balance for the electron movement in the direction parallel to E\mathbf{E} and perpendicular to E\mathbf{E} reads

ρeE=12dεdxE2;(85)\rho_e E = \frac{1}{2}\frac{\mathrm{d}\varepsilon}{\mathrm{d}x}E^2; \tag*{(85)}
JB=12d(1/χ)dxB2.(86)JB = \frac{1}{2}\frac{\mathrm{d}(1/\chi)}{\mathrm{d}x}B^2. \tag*{(86)}

Combing these two equations, we obtain the following condition for the movement of electrons

Jρe=d(1/χ)dεBE⇔Jρe=C2EB,(87)\frac{J}{\rho_e}=\frac{\mathrm{d}(1/\chi)}{\mathrm{d}\varepsilon}\frac{B}{E}\Leftrightarrow\frac{J}{\rho_e}=\frac{C^2}{E}B, \tag*{(87)}

where C=1/(εχ)C=\sqrt{1/(\varepsilon\chi)} denotes the speed of light in media. Due to axi-symmetry, the vectors, E\mathbf{E}, B\mathbf{B}, and J\mathbf{J} are represented by the scalar fields EE, BB, and JJ, respectively.

From Eq. (87), the electrical resistance rr at any position is expressed as

r=EJ=ρeC2J2B.(88)r=\frac{E}{J}=\frac{\rho_e C^2}{J^2}B. \tag*{(88)}

When B=0B=0, there is no resistance and we have superconducting. The superconducting with B=0B=0 is consistent with the Meissner effect [12, 13].

Speed of light

In vacuum, we have ρe=0\rho_e=0 and therefore J=0\mathbf{J}=0. By considering these two conditions, the Maxwell-Ampere equation is modified as

χ0ε0dEdt=∇×B,(89)\chi_0\varepsilon_0\frac{\mathrm{d}\mathbf{E}}{\mathrm{d}t}=\nabla\times\mathbf{B}, \tag*{(89)}

where χ0\chi_0 and ε0\varepsilon_0 are the magnetic permeability and electrical permittivity, respectively, in vacuum. By taking ∇×\nabla\times on both sides of Eq. (89), we obtain

χ0ε0d2Bdt2=∇2B.(90)\chi_0\varepsilon_0\frac{\mathrm{d}^2\mathbf{B}}{\mathrm{d}t^2}=\nabla^2\mathbf{B}. \tag*{(90)}

By using the total derivative for a one-dimensional setup without loss of generality in xx-coordinate,

dBdt=∂B∂t+dBdxdxdt=∂B∂t+vdBdx,(91)\frac{\mathrm{d}B}{\mathrm{d}t}=\frac{\partial B}{\partial t}+\frac{\mathrm{d}B}{\mathrm{d}x}\frac{\mathrm{d}x}{\mathrm{d}t}=\frac{\partial B}{\partial t}+v\frac{\mathrm{d}B}{\mathrm{d}x}, \tag*{(91)}

we obtain the second derivative of the magnetic field as

d2Bdt2=∂2B∂t2+dvdtdBdx+2v∂∂tdBdx+v2d2Bdx2.(92)\frac{\mathrm{d}^2B}{\mathrm{d}t^2}=\frac{\partial^2B}{\partial t^2}+\frac{\mathrm{d}v}{\mathrm{d}t}\frac{\mathrm{d}B}{\mathrm{d}x}+2v\frac{\partial}{\partial t}\frac{\mathrm{d}B}{\mathrm{d}x}+v^2\frac{\mathrm{d}^2B}{\mathrm{d}x^2}. \tag*{(92)}

For the steady state, ∂X/∂t=0\partial X/\partial t=0, X=BX=B, dB/dx\mathrm{d}B/\mathrm{d}x and assuming that there is no acceleration, dv/dt=0\mathrm{d}v/\mathrm{d}t=0, we obtain d2B/dt2=v2d2B/dx2\mathrm{d}^2B/\mathrm{d}t^2=v^2\mathrm{d}^2B/\mathrm{d}x^2 and hence the speed of light in vacuum as

v2=C2=1χ0ε0,orC=1χ0ε0.(93)v^2=C^2=\frac{1}{\chi_0\varepsilon_0},\quad\text{or}\quad C=\sqrt{\frac{1}{\chi_0\varepsilon_0}}. \tag*{(93)}

Similarly, we can also derive the speed of light in media.

We propose a more general formulation to define the speed of light. At the steady state, the momentum balance equation reduces to

v⋅∇(B×D)=−∇p2(94)\mathbf{v} \cdot\nabla(\mathbf{B} \times\mathbf{D}) = -\nabla p_{2} \tag*{(94)}

Similar to the treatment of mass momentum, we assume the electromagnetic momentum takes the form

B×D=ρpv,(95)\mathbf{B} \times\mathbf{D} = \rho_{p}\mathbf{v}, \tag*{(95)}

where ρp\rho_{p} represents an effective “mass density” of the electromagnetic field, for instance, associated with photons. Under this assumption, we obtain a general expression for the speed of light that is directly analogous to the formulation of the speed of sound:

v=dp2dρp.(96)v = \sqrt{\frac{d p_{2}}{d \rho_{p}}}. \tag*{(96)}

Sound speed

Neglecting the Korteweg stress tensor in the momentum conservation of mass leads to

ρdudt+udρdt=−∇p1.(97)\rho\frac{d\mathbf{u}}{dt} + \mathbf{u}\frac{d\rho}{dt} = -\nabla p_{1}. \tag*{(97)}

At isentropic condition implying that the time scale of dsu/dtd s_{u}/dt is much less than dρ/dtd\rho/dt, we have

dudt=0.(98)\frac{d\mathbf{u}}{dt} = 0. \tag*{(98)}

Substituting Eq. (98) into Eq. (97) and considering the steady state in 1D, we obtain

u2dρ=−dp1.(99)u^{2}d\rho= -d p_{1}. \tag*{(99)}

With the expression for the pressure −p1=f−μc≔P-p_{1} = f - \mu c \coloneqq P, we obtain the formulation for the speed of sound as

u=dPdρ.(100)u = \sqrt{\frac{dP}{d\rho}}. \tag*{(100)}

Parity asymmetry

Parity (mirror) symmetry refers to invariance under spatial inversion, x→−x\mathbf{x} \to-\mathbf{x} [3]. A system preserves parity if its equations of motion remain unchanged under this reflection operation. If the physical force or field reverses sign under the parity transformation, the system exhibits parity violation or broken mirror symmetry. In the classic concept [14], B\mathbf{B} and H\mathbf{H} are even parity with B(−x)=B(x)\mathbf{B}(-\mathbf{x}) = \mathbf{B}(\mathbf{x}); D\mathbf{D} and E\mathbf{E} are odd parity with D(−x)=−D(x)\mathbf{D}(-\mathbf{x}) = -\mathbf{D}(\mathbf{x}). This concept is mainly based on E=∇ψ0\mathbf{E} = \nabla\psi_0 and H=∇×φ1\mathbf{H} = \nabla\times\boldsymbol{\varphi}_1. As ∇−x=−∇x\nabla_{-\mathbf{x}} = -\nabla_{\mathbf{x}}, we have E(−x)=−E(x)\mathbf{E}(-\mathbf{x}) = -\mathbf{E}(\mathbf{x}). As φ1(−x)=−φ1(x)\boldsymbol{\varphi}_1(-\mathbf{x}) = -\boldsymbol{\varphi}_1(\mathbf{x}), we have H(−x)=H(x)\mathbf{H}(-\mathbf{x}) = \mathbf{H}(\mathbf{x}).

From the momentum evolution of the electromagnetic field, we obtain a generalized electromagnetic force

f=−∇⋅Γ+ρe(∇×ψ1);(101)\mathbf{f} = -\nabla\cdot\boldsymbol{\Gamma} + \rho_e(\nabla\times\boldsymbol{\psi}_1); \tag*{(101)}
Γ=−(12D⋅E+12B⋅H)I+D⊗E+B⊗H.(102)\boldsymbol{\Gamma} = -\left(\frac{1}{2}\mathbf{D}\cdot\mathbf{E}+\frac{1}{2}\mathbf{B}\cdot\mathbf{H}\right)\mathbf{I}+\mathbf{D}\otimes\mathbf{E}+\mathbf{B}\otimes\mathbf{H}. \tag*{(102)}

The first force term in Eq. (101) is contributed by the classic Maxwell stress tensor. This classic term gives rise to the symmetry of the parity. Under a mirror (parity) transformation, D\mathbf{D} and E\mathbf{E} both change sign, becoming −D-\mathbf{D} and −E-\mathbf{E}. However, the sign of D⊗E\mathbf{D}\otimes\mathbf{E} and D⋅E\mathbf{D}\cdot\mathbf{E} remains invariant under this transformation. The momentum balance equation after applying the mirror operation becomes

−d(B×D)dt=∇⋅Γ.(103)-\frac{\mathrm{d}(\mathbf{B}\times\mathbf{D})}{\mathrm{d}t}=\nabla\cdot\boldsymbol{\Gamma}. \tag*{(103)}

The negative sign on the left-hand side arises from the sign reversal of D\mathbf{D} under the mirror transformation. The electromagnetic force changes from −∇⋅Γ-\nabla\cdot\boldsymbol{\Gamma} to ∇⋅Γ\nabla\cdot\boldsymbol{\Gamma} is due to ∇−x=−∇x\nabla_{-\mathbf{x}}=-\nabla_{\mathbf{x}}. As demonstrated in Eq. (103), the mirror operation does not change the momentum balance equation.

The second term in Eq. (101) leads to the breaking of parity symmetry. In the mirror transformation, the sign of ∇×ψ1\nabla\times\boldsymbol{\psi}_1 is unchanged, as ∇−x=−∇x\nabla_{-\mathbf{x}}=-\nabla_{\mathbf{x}} leads to a negative sign and ψ1\boldsymbol{\psi}_1 becomes −ψ1-\boldsymbol{\psi}_1 (the vector ψ1\boldsymbol{\psi}_1 is an odd parity). This results in the non-conservation of parity. In fact, when E=∇ψ0+∇×ψ1\mathbf{E}=\nabla\psi_0+\nabla\times\boldsymbol{\psi}_1, the Maxwell stress tensor is not mirror symmetric anymore due to the cross term, (∇ψ0)⊗(∇×ψ1)(\nabla\psi_0)\otimes(\nabla\times\boldsymbol{\psi}_1), which also leads to non-conservation of parity.

Conclusion

In summary, we have developed a generalized theory for the electromagnetic field based on the simple yet fundamental principle of entropy maximization, offering an alternative interpretation of entropy within the system. The proposed framework has been validated by its ability to reproduce the classical Maxwell equations, the propagation speeds of light and sound, and the Meissner effect in superconductors. A key outcome of the theory is a generalized electromagnetic force that includes not only the classical Maxwell stress tensor but also an additional term, ρe(∇×ψ1)\rho_e(\nabla\times\psi_1), arising from the motion of charges. This new contribution provides a natural explanation for the non-conservation of parity in electromagnetic interactions.

Declaration of competing interest

The author declares that he has no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

No data was used for the research described in the article.

References

  1. [1]M. J. Benacquista and J. D. Romano, Classical mechanics (Springer, 2018).
  2. [2]J. S. Rowlinson, Journal of Statistical Physics 20, 197 (1979).
  3. [3]L. D. Landau, J. S. Bell, M. Kearsley, L. Pitaevskii, E. Lifshitz, and J. Sykes, Electrodynamics of continuous media, Vol. 8 (elsevier, 2013).
  4. [4]F. Wang and B. Nestler, arXiv preprint arXiv:2406.19786 (2024).DOI
  5. [5]Z. L. Wang, Materials Today 52, 348 (2022).
  6. [6]J. W. Cahn and J. E. Hilliard, The Journal of chemical physics 28, 258 (1958).
  7. [7]Y. Cai, F. Wang, H. Zhang, and B. Nestler, Journal of Physics: Condensed Matter 36, 495702 (2024).
  8. [8]K. S. Ridl and A. J. Wagner, Physical Review E 98, 043305 (2018).
  9. [9]W. Sprössig, Mathematical methods in the applied sciences 33, 374 (2010).
  10. [10]H. Zhang, F. Wang, and B. Nestler, Journal of Computational Physics 505, 112907 (2024).
  11. [11]S. Chen and G. D. Doolen, Annual review of fluid mechanics 30, 329 (1998).
  12. [12]W. Meissner and R. Ochsenfeld, Naturwissenschaften 21, 787 (1933).
  13. [13]N. Gavish, O. Kenneth, and A. Keren, Physica D: Nonlinear Phenomena 415, 132767 (2021).
  14. [14]L. D. Landau, E. M. Lifshitz, and E. Lifshitz, The Classical Theory of Fields: Volume 2.

Paper details

vlnc:W_ny8s9dk4
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents